We previously evaluated an attenuated chimeric human/bovine PIV3 as a vector to express the spike glycoprotein of Severe Acute Respiratory Syndrome Coronavirus (SARS) from an added gene. A single dose of this construct administered to the respiratory tract was immunogenic and protective against SARS challenge in African green monkeys (AGM). We also evaluated HPIV3 expressing the single glycoprotein GP of Ebola virus Zaire. Wild type HPIV3 is naturally attenuated in non-human primates and thus is an appropriate model for an attenuated PIV vector. In guinea pigs, a single respiratory tract inoculation was completely protective against an otherwise highly lethal intraperitoneal challenge of Ebola virus Zaire. In rhesus monkeys, a single immunization protected 88% of animals against an otherwise highly lethal intraperitoneal challenge, and two doses provided complete protection. Thus, immunization via the respiratory tract indeed was substantially protective against parenteral infection. We also showed that HPIV3 expressing Ebola GP was comparably immunogenic in HPIV3-seropositive monkeys following two doses. This showed that the vaccine should be effective in vector-immune animals. We also showed that the HN and F surface antigen genes of the HPIV3 vector could be replaced by the GP gene of Ebola virus, resulting in a highly attenuated virus in which the only viral surface protein was GP of Ebola. When evaluated in guinea pigs, a single dose was fully protective against challenge with guinea pig-adapted Ebola virus. Further studies are planned. We also have been investigating the use of NDV as a human vaccine vector. NDV is antigenically distinct from common human viruses and should not be affected by pre-existing immunity to common viruses. There is anecdotal evidence that NDV is highly restricted in humans and does not cause significant disease. Previously, we confirmed that NDV is very highly restricted following combined intranasal and intratracheal inoculation of rhesus monkeys and AGM;indeed, NDV replication could barely be detected. Nonetheless, we found that expressed foreign proteins are surprisingly immunogenic and protective. We previously showed that immunization of AGM with two doses of NDV expressing the SARS S protein was strongly immunogenic and protective against challenge with a high dose of SARS. Similarly, immunization of AGMs with two doses of NDV expressing either the hemagglutinin (HA) or neuraminidase (NA) glycoprotein of highly pathogenic avian influenza virus (HPAIV) H5N1 induced high levels of serum neutralizing antibodies and complete or nearly-complete protection against replication of a high dose of challenge HPAIV. In the present reporting period, we extended these studies by evaluating NDV that was engineered to express the GP of Ebola virus Zaire. In rhesus monkeys, a single dose of NDV-GP induced low levels of GP-specific antibodies in serum and respiratory tract secretions compared to HPIV3 expressing GP (used as a positive control). Following a second dose of NDV-GP, titers of GP-specific serum antibodies detected by ELISA remained somewhat lower than those induced by two doses of HPIV3-GP, but the titers of GP-specific secretory antibodies and serum neutralizing antibodies induced by the different vectors (NDV versus HPIV3) were indistinguishable. To investigate the range of usefulness of the NDV vector, we also expressed envelope glycoprotein gp160 of human immunodeficiency virus 1 (HIV-1) from an added gene. Guinea pigs were immunized and boosted by various combinations of the intranasal and intramuscular routes, resulting in HIV-1-neutralizing antibody responses. The intranasal route of immunization and boosting proved to be more immunogenic than the intramuscular route and induced gp160-specific antibody responses detected both in sera and vaginal washes. NDV vectors also were constructed to express either of two antigens for the Lyme disease pathogen Borrelia burgdorferi. These vectors were poorly immunogenic in mice but were substantially immunogenic in hamsters. Immunization of hamsters with NDV expressing one of the antigens, BmpA (expressed either as a transmembrane or secreted form), was associated with a partial reduction in pathogen load. It should be noted that Borrelia is refractory to immunoprophylaxis in general. This showed that NDV vectors can induce immunity against a bacterial pathogen, although the level of protection was low. However, the main use of NDV as a vector would be against viral pathogens, particularly those that can use respiratory and conjunctival membranes as portals of entry. NDV represents serotype 1 of the APMVs. There are 8 other serotypes, namely APMV2 to 9. There is a great deal of information available for NDV because of its importance in poultry. In contrast, relatively little was known about the other serotypes, apart from a complete sequence for one strain of APMV-6. Working with collaborators at the University of Maryland at College Park, we previously initiated antigenic and sequence analysis of these as a prelude to their evaluation for safety, replication, and immunogenicity in non-human primates as potential vectors. In studies continuing in the present year, we have determined complete consensus sequences for all of the serotypes, and in a number of instances have sequences for two to four different strains for each serotype. This has provided evidence for antigenic and sequence heterogeneity within serotypes. For example, examination of several strains of APMV2 this year provided genetic evidence of two subgroups. These also were shown to be distinguishable by reactivity with convalescent antisera. Similarly, analysis of several APMV6 strains this year provided evidence of sequence and antigenic dimorphism. In addition, reverse genetics systems have been produced for a number of serotypes, including serotypes 2, 3, 4, and 7. In NDV and avian influenza viruses, the presence of a multi-basic cleavage sequence in the F (NDV) or HA (AIV) protein is well-known to be associated with efficient cleavage and a pantropic (and virulent) phenotype, whereas the presence of fewer basic residues is associated with a need for added trypsin for replication in vitro and a lack of systemic spread in vivo (and an avirulent phenotype). Unexpectedly, for a number of the other APMV serotypes, this association was not observed. For example, efficient cleavage of the F protein often did not depend on a multi-basic cleavage sequence. Also, studies with APMV2 showed that the virus spread systemically in chickens even though it lacked a multi-basic cleavage site. Furthermore, even though APMV2 spread systemically and produced large amounts of antigen, infection was asymptomatic. APMV2 was modified so that its F protein cleavage site (which is not multi-basic) was replaced with 11 different cleavage sites, including a number of multi-basic sites. Each mutant exhibited a lack of dependence on added trypsin for cleavage (like the APMV2 parent), and each mutant gained the ability to form syncytia in cell culture (whereas the APMV2 parent does not form syncytia). Unexpectedly, all of the mutants grew at least 10-fold more efficiently in vitro than the APMV2 parent. Examination of selected viruses confirmed increased cleavage of the F protein. Despite the gains in syncytium formation, replication, and F protein cleavage, examination of selected mutants in chicken eggs and 1-day-old and 2-week old chickens showed that they fully retained the avirulent phenotype of the APMV2 parent. Thus, the canonical link between multi-basic cleavage site, syncytium formation, systemic spread, and virulence was not observed. In addition, this showed that mutants with improved growth could be developed without concomitant increases in virulence, which may be useful for vaccine purposes.